The Fluoride Complexing of Scandium(III) in Aqueous Solution: Free

John W. Kury, Armine D. Paul, Loren G. Hepler, and Robert E. Connick. J. Am. Chem. Soc. , 1959, 81 (16), pp 4185–4189. DOI: 10.1021/ja01525a018...
0 downloads 0 Views 598KB Size
COa and CaO were. However, the sample which was treated in an identical manner, except t h a t it was not exposed to the atmosphere, showed no evidence of reaction. The original BaO and CaC03 were still present despite t h a t fact t h a t the sample had been heated for 20 minutes at 125’ above the “reaction temperature.” This same measurement has been made with BaO CuSO4, SrO CaC03 and SrO CuSO4 with results the same as described. Specifically, no reaction takes place in any of these other three systems when the reactants are prepared, mixed and heated above the “reaction temperature’’ in BUCUO. Under conditions which are otherwise identical, the control which is exposed to the atmosphere does react in each case. It appears, therefore, t h a t the reactions of BaO and SrO as described by Hedvall do not proceed by solid state diffusion processes but occur via the liyuid phase. Furthermore, the so-called characteristic “reaction temperature” has nothing to do with the oxide itself but is merely the melting point of what is probably a hydroxide containing eutectic as Balarew suggested,s the presence of which presents a low energy path along which the reaction proceeds very rapidly. This is illustrated in Fig. 2 where hypothetical rate constant temperature curves are shown for the liquid phase and solid state mechanisms by eurves I and 11, respectively. Below the melting point of the third component ( T M )the , only path available is the solid state path14 where, even

+

+

I

/ El / /

/

/ /

+

(14) b’e can, of course, o n l y speculate as t o t h e reaction p a t h in t h e absence of a liquid phase a n d a better guess, probably, is t h a t t h e reac-

/

/

(M

TEMPERATURE

-

‘C.

Fig. 2.-Postulated inccliariisin c ~ f the reactir~iis i l f BaO and SrO. Hypotlictical rate cuiiituiit tcinpcrnture curves for the liquid phase anti solid st:ttc meclianistns are slir~wii, respectively, as cur\’cs I anid 11. Tlic rcacticm patli (svlid line) follows curvc 11 up to the iiieltiiig point of the third conipoiieiit at which piiiiit tlic lowcr energy, liquid phase path becomes available.

a t Tiif, the reaction rate is so small t h a t it may be considered to be zero for all practical purposes. On sudden appearance of a liquid phase, the system finds itself well up on the steep portion of curve I where. because of its exotherniic nature. the reaction proceeds with a burst of heat, leading to the high AT’S observed with DTA. tion will proceed through t h e gaseous phase. pose of t h e above, t h i s is of no consequence.

IIowever, for thc p u r -

SCIIESECTADP, ru’S\V T O R K

[COSTRIBLJTION FROM T H E DEPARTMENT O F CIIEMISTRY A S D T H E R A D r a r I o N LABORA4TORYO F THE USIVERSITV O F CALIFORSIA, BERKELEY]

The Fluoride Complexing of Scandium(II1) in Aqueous Solution : Free Energies, Heats and Entropies BY JOHN U‘.KURY,ARMINED . PAUL,LORENG. HEPLERAXD ROBERTE. CONNICK RECEIVED FEBRUARY 6, 1959 T h e complexing of Sc(II1) in perchlorate solutions at an ionic strength of 0.5 M w t s deterrnined b>-studying the potential of the ferrous-ferric electrode as a function of the HF, H + and Sc(II1) concentrations. The data can lie interpreted by assuming the presence of the species ScF++, ScF*+,ScFa(aq) and SCFI-. Equilibrium constants for the fortnation of these complexes at 15, 25 and 35” were evaluated and from them the heat and entropy changes of the reactions calculated. The scandium fluoride complexes are more stable than those of other trivalent ions of comparable radius. The extra stability of the complex scandium fluorides arises from an irregularity in the heat term, rather than in the entropy term.

Introduction scandium fluoride complexes arises from the heat This investigation was undertaken because of the term or the entropy term. The “ferri” method of Brosset a i d Orring4 was work of Dodgenl who found the scandium fluoride complexes t o be more stable than those of other used. The complexing of Sc(II1) was deterniinetl trivalent ions of similar radius. According oto indirectly by studying its effect on the ferric fluoride Pading,‘ In(II1) has the same ionic radius (0.81 A.) complexing equilibria, as measured potentiometrias Sc(II1). However, the first fluoride complex of cally through the ferrous-ferric couple. Sc(II1) is about 250 times more stable than the Experimental first fluoride coniplex of I n ( I I I ) . 3 \Ve wished to Apparatus.~-L)cccripti(iniof the cclli, clcctrotlci, incaisurdetermine whether the greater stability of the liig Lil)p.tr.itui c i i i d geueral procidurc x e gib cii clie\\licrc..3 ‘M (1) H. W. Dodgen, private communicatiun. ( 2 ) L. Pauling, “ T h e Piature of t h e Chemical Bond,” Cornell University Press, I t h a c a , N . Y . , 1945. (3) L. G . Hepler, J. W.Kury and Z . Z . Hugus, J . P h v s Chci~z.,68, 2(i (1951).

(4) C . Brosset and J. Orring, Saeusk. Keiri. T i d , 66, 101 (1943). ( 5 ) H . I’J Dodgen and G. K . Rollelson, TEXIS J O U R N A L . 71, l’IiO0

(lQ49). ( t i ) I2 E Connick a n d R I . Tbao, i h i d . . 7 6 , ::!I1

(1051)

Solutions.-A scandium perclilorate-perchloric acid stock solution was prepared by dissolving spectroscopically pure S c ~ 0 3in perchloric acid. The solution was analyzed for scandium by precipitating it as the hydroxide and weighing as Sc203. The acidity was determined by precipitating the scandium as ScFa and titrating with standard base t o the phenolphthalein end-point. Double vacuum-distilled perchloric acid was diluted with water and the solution standardized against mercuric oxide. Sodium fluoride solutions were prepared by dissolving weighed amounts of the C.P. salt which had been dried a t 150°. A stock solution of sodium perchlorate was prepared by treating C.P. sodium carbonate with a n excess of perchloric acid. After boiling t o cxpel the carbon dioxide, the pH was adjustcd to 5-6 with carbonate-free sodium hydroxidc and thc solution then filtered. Aliquots were evaporated to dryness, dried for several hours a t 160", and weighed as SaClOa. -1 ferrous perchlorate-ferric perchlorate-perchloric acid stock solution W:LS prepared by dissolving a weighed amount of electrolytic iron it1 a known amount of perchloric acid. 4 portion of the resulting fcrrous perchlorate was oxidized to ferric perchlorate with h~-drogenpcroxide. The ferrous ion coricciitratiuii was detcrrnined by titration with standard ccric sulfatc. The ferric ion concentration and thc acidity ivcre determined by difference. Distilled water was redistilled from alkaline permaiigaiiate for use in all solutions. Procedure .-Three half-cells were used wliicli will bc tlesignated A, B and C. Half-cells A and C, and half-cells B and C , were connected by sodium perchloratc-agar agar salt bridgcs. Each half-cell initially contained the same concentrations of ferrous perchlorate, ferric pcrchloratc and perchloric acid. Half-cell A contained in addition a known concentration of scandium perchlorate. The ionic streiigtli in each half-cell was adjusted to 0.50 :I1 with sodiuin perclilorate. The volumes were equal: After the initial zero potentials 111 each ccll became COIIs t a n t , known quantities of sodiuin fluoride wcre added to half-cells h and B from a calibrated micropipet, and the potential measured after cach xdditioii. Thc fluoride additioii was coiitinuctl until a potential of xbout 150 mv. w:is developed iii each ccll or uiitil scandium fluoridc precipit:itctl. The potential readiiig bccmie constant to i i few huntlredtiib of a millivolt within about oiic iniiiutc after eacli fluoride addition.

Experimental Data The initial conditions for each experinleiit are summarized in Table I. TABLE I I N I T I A L E X P E R I M E N T A L CONDITIONS Sc(C104)3 , P e ( C I O h , I'e(ClOa)a, moles/l. IIClOi, moler/l. Temp., moles/l. x 104 x in4 x 108 moles:). OC.

25 25 25 35 35 35 15 15

15

73 387 735 387 741 6 665 1 387 6 665 2 741

12 1 2 1 2

0.1122 ,05287 ,05453 ,05277 ,05493 lOfi6 ,053'77 . l06,j

. U54!)3

3 3 3 3 :3

553 869 831 868 831 :3 '721 :j 86X :3 721 :3 831

3 3 3 3 3 3 3 3 :3

NaClOi molesfl:

402 705 669 704 008 56:; 704 563

668

The acidity was varied twofold aiid the scaiidiuiii concentration was varied a t least fivefold. The maximum stoichiometric fluoride concentration depending on the varied from 0.015 to 0.03 particular experiment. Sample data for the second experiment of Table I are given in Table 11. Data for all b u t the last experiment are given by P a d 8 ( 7 ) T h e symbol M indicates concentrations in moles per liter of solution. (8) A. I). Paul, 'Thesis, University (if California, 19.55; 7Jniver4ty of California Radiation Laboratory unclasbified report 7JCRI. 2!)2fi,

?'AIII,E I1 DATAIWK SECOND EXPERIMENT 01' TAHLE I

0.4702 JM Sal' added Cell A-C ( S c + + +hy resent i t ] A) N a F added,ml. E (mv.)

I n i t i a l vol. = 100.05 ml. Cell D-C (no S c + present) Nal; a d d e d , mi. E (rnv.) + +

0 0747

13 24 34 42 40 55 61 66 70 93 110 121 136 14(i

1401 2241 2988 3735

1482

1 1 3 2 3

3229 5976, 672.3 1707 0091 1675 6659 16 13

25

0 0747

0 . 72

1494 224 1 29x8 4442 59'70 8961

3.16 5.12 11.15 21). 1s 'I 1 . n!)

98 71 84

75 84 24 11

3'; 07 77 57 IO 80

1.74

I (14%

51.80

1 1952 1 5687 2 l4lh 2 6402 3 13x0 3 0370

t i 0 . ($0 TX . (i8

!)$I. 97 114.89

127.61 lXJ. 00

Analysis of Data 'I'he equilibria in question are

+ + +

+

S C + ~ I1F = SCE'+~ E1 i. SCF+2 € I F = ScI'g+ H' ScF-+ €IF = scli3('Lq) 13 S C F ~ ( ' L ~ )HI' = SCFI- H c = I'cp+2 + H + I>e+3 4Fc17 + 2 E l l = FCFZ+ II I'CFLt IIF = FCF3(:1q) 11 Fc-3 1 p = 1;eOI-I + 2 11' IIF = I I + 1'-

+ + +

+

+

+ ' +

+ ' + +

+

l'he hydrolysis of scandium ion9 a i d the hydrolytic diirierizatiori of ferric ioii'O were both negligible under the conditions criiployed here. The Q ' s arc cquilibriuiii quotients expressed iii coiicentratioiis, for exainplc where the quantities iii parciithcses represent the concentrations of the eiiclosed species in moles per liter. The values a t 1 = 0.5 of Q6, Q 6 , Q 7 , Q h and @IF used in these calculations are given in ref. 11. The difference between the moles of fluoride added to half-cell A and the moles of fluoride added to half-ccll B when the potentials are equal is, to a very good approximation, the moles of fluoride iii half-cell X present in the form of scandium fluoride complexes. Division by the volume of solution of half-cell A yields the concentration of fluoride as scandium complexes, symbolized by A ( S F - ) . The moles of fluoride in half-cell I? corrcspoiidiiig to each potential reading of A were read from a plot of the former wrszds potential. From stoichiometry i t can be seen t h a t A(zF-) (ZSC+3) -

+ 3(SCl'.i) + 4(Scl'r ) (11) = (SCtJ) + (ScF'.Z) + (SCl'l+) + (%Fa) + (SCF~-)( I 2 ) = (SCF+2)

+ 2(

j J ~ )

n

April, 1955. T h e values listed for are slightly in error a t high fluoride concentrations because of t h e use of a poorer approximation t h a n used here in calculating